The One-Line Fix That Stops AI Training On Your Site

OpenAI announced on 30 October 2026 that it will honour a new shared robots.txt signal, giving website operators a single way to express their AI training preferences. Google and Microsoft are backing the same standard, meaning one line added to a website's robots.txt file can opt that site out of AI training across several major providers at once, rather than listing separate rules for every crawler.

What it means for your hours and your content

If you run a trade or construction business with a website showing project case studies, before-and-after photos, pricing guides or technical write-ups, that content has likely already been scraped by AI crawlers to train models. Once trained into a model, your unique content, your way of explaining a job, your pricing logic, can effectively be reused by anyone asking an AI chatbot a similar question, including your competitors.

Worked example: a kitchen fitting firm has a blog of 40 detailed project write-ups built up over three years, each one a selling point that took real time to produce. Previously, blocking AI training meant a web developer manually adding exclusion rules for each AI company's specific crawler, GPTBot, Google-Extended and others, a fiddly job that is easy to get wrong or forget to update as new crawlers appear. With the new shared signal, that same protection is one line, checked once and done.

This will not get your business more leads this week. What it does is close a quiet leak, your competitive content being absorbed into tools your competitors can also use, for free, without your name attached. For a small business whose website content is a genuine differentiator, that is worth five minutes of a developer's time.

What to do this month

• Ask your web developer or hosting provider whether your current robots.txt file already blocks AI crawlers, and how.
• Request the new shared opt-out signal be added, covering OpenAI, Google and Microsoft in one line rather than separate rules.
• Decide, content by content, whether you actually want to block training. Some owners are happy for general blog content to be found by AI search tools but want project pricing and client details excluded.
• Put a reminder in six months to check whether other AI providers have joined the standard, so your opt-out stays current.